作者: Jhun-Ying Yang , Yen-Ping Chen , Gwo-Yun Lee , Shun-Nan Liou , Jeen-Shing Wang
DOI: 10.1007/978-3-540-74873-1_47
关键词:
摘要: This paper presents a neuro-fuzzy classifer for activity recognition using one triaxial accelerometer and feature reduction approaches. We use to acquire subjects' acceleration data train the neurofuzzy classifier distinguish different activities/movements. To construct classifier, modified mapping-constrained agglomerative clustering algorithm is devised reveal compact configuration from data. In addition, we investigate two methods, subset selection linear discriminate analysis. These methods are used determine significant subsets retain characteristics of distribution in space training classifier. Experimental results have successfully validated effectiveness proposed